• Citer

Evaluation of multi-platform measurements of tropospheric ozone precursors using a regional model

par Poraicu, Catalina
Président du jury Coheur, Pierre
Promoteur Clarisse, Lieven
Co-Promoteur Muller, Jean François
Publication Non publié, 2025-09-04

Thèse de doctorat

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

Despite their trace concentrations, nitrogen oxides (NOx = NO + NO2) and volatile organic compounds (VOCs) are central to atmospheric reactivity. As precursors of tropospheric ozone and fine particulate matter, they contribute to air quality degradation and climate change. In addition, NOx and VOCs affect the oxidizing capacity of the atmosphere by modulating levels of the hydroxyl radical (OH). However, the sources and fate of NOx and VOCs remain poorly quantified, especially in regions where observational coverage is limited.This thesis uses the high-resolution atmospheric regional model WRF-Chem (1) to better understand the budgets and transformations of tropospheric NOx and VOCs, in combination with multiple observational datasets, (2) to validate and characterize high-resolution spaceborne measurements of tropospheric NO2 columns, (3) to demonstrate the value of satellite data for constraining emissions in both polluted and remote environments, and (4) to assess the sensitivity of oxidant levels (OH, O3) to changes in emissions in polluted and remote environments. High resolution modelling is particularly useful for capturing the spatiotemporal variability of reactive species and is therefore best suited to exploit field composition measurements in regions with strong horizontal gradients such as cities and industrial areas. A first case study focuses on the region of Antwerp, Belgium, an area strongly influenced by anthropogenic emission sources. The model framework, constrained by observations, is used to quantify biases in satellite NO2 retrievals with respect to airborne remote sensing data. In combination with satellite and in situ data, it is also used to improve the temporal profile and spatial distribution of NOx emissions in the model. This studydemonstrates how models can complement validation methods by addressing differences in sampling time, vertical sensitivity, and spatial coverage. A second case study addresses the remote tropical site of Réunion Island, featuring low NOx levels and a complex mix of natural and anthropogenic emissions. Comparisons with satellite and in situ data suggest overestimations of industrial NOx and residential VOC emissions, as well as severe underestimations of natural sources of key organic compounds, either of marine or biogenic origin. The study also points to limitations in the currentrepresentation of VOC emissions and chemistry in the WRF-Chem model, which should be further investigated in future studies.Finally, a comparative assessment of the sensitivity of ozone and radicals to changes in precursor emissions shows very different chemical regimes in the two regions: NOx-saturated over Flanders, where elevated NOx levels suppress oxidant levels, and NOx-limited in Réunion, where oxidant recycling is constrained by NO availability. These findings underscore the importance of accounting for local chemistry when designing effective air quality strategies.